Self-selected walking speed is a key outcome for exercise-based rehabilitation programs following lower-extremity trauma. This work introduces a novel reinforcement learning-based assist-as-needed (RL-AAN) controller for ankle exoskeletons, aimed at gait speed training. Built on an actor–critic architecture, the RL-AAN controller integrates a control objective that balances the trade-off between expected stride velocity (SV) errors and exoskeleton assistance. This approach allows the exoskeleton to progressively reduce ankle plantar- and dorsiflexion (PDF) assistance as the user’s performance improves, promoting active participation. The desired assistive torque is computed as the product of the actor output and the wearer’s biomechanical ankle PDF moment, estimated by a subject-agnostic model, thereby ensuring personalized and biomechanically relevant assistance. In a proof-of-concept study with healthy individuals walking on a self-paced treadmill with ankle weights, the RL-AAN controller outperformed a conventional Fixed-K controller—achieving greater immediate speed increases during assisted walking (14.2% vs. 10.0% relative to unassisted perturbed walking) and inducing short-term gait speed adaptation post-training, not observed with the conventional controller. These findings highlight the potential of RL-AAN control for subject-tailored gait training, with promising clinical implications for exercise-based rehabilitation in individuals with neurological or musculoskeletal gait impairments.

Research in powered prostheses and orthoses has relied on COP measurements to inform a device’s controller about the body’s progression through the gait cycle, and to provide sensory substitution for prosthesis users, thereby helping them maintain balance during locomotion. Obtaining accurate COP measurements in out-of-the-lab contexts currently requires pressure sensitive insoles with dense arrays of sensing elements, which are expensive and bulky, limiting the accessibility and scalability of this technology. In this paper, we present a new method to reconstruct COP trajectories in over-ground walking tasks, using an affordable sensor array with eight sensing elements embedded in shoe insoles. The method leverages Gaussian Process Regression (GPR) to perform predictions from raw sensor data using Bayesian inference. A preliminary validation was carried out with a convenience sample of healthy individuals and patients with neuromuscular disorders. Combined mediolateral (ML) and anteroposterior (AP) errors where 2% and 3% for healthy individuals and patients, respectively. The analysis evidenced larger stride-to-stride variability in the ML COP excursion for the patient group, suggesting higher levels of motor noise associated with selective muscle weakness. These promising results indicate the potential of the proposed method to accurately estimate COP trajectories for future applications in wearable robotics and out-of-the-lab clinical gait assessments.

Online estimation of 3D ground reaction forces (GRFs) is becoming increasingly important for closed-loop control of lower-extremity robotic exoskeletons. Through in-verse dynamics and optimization models, 3D GRFs can be used to estimate net joint torques and approximate muscle forces. Although instrumented footwear to measure vertical GRFs in out-of-the-lab environments is available, accurately measuring shear GRFs with foot-mounted sensors still remains a challenging task. In this paper, a new outsole-embedded optoelectronic sensor configuration that is able to measure biaxial shear GRFs is proposed. Compared with traditional strain-gauge based solutions, optoelectronic sensors allow for a more affordable design. To mitigate the risk of altering the wearer's natural gait, the proposed solution does not involve external modifications to the footwear structure. A preliminary validation of the outsole-embedded sensor was conducted against validated laboratory equipment. The test involved two sessions of treadmill walking at different speeds. Experimental results suggest that the proposed design may be a promising solution for measuring shear GRFs in unconstrained environments.

In this paper, we propose an autonomous gait analysis system consisting of a mobile robot and custom-engineered instrumented insoles. The robot is equipped with an on-board RGB-D sensor, the insoles feature inertial sensors and force sensitive resistors. This system is motivated by the need for a robot companion to engage older adults in walking exercises. Support vector regression (SVR) models were developed to extract accurate estimates of fundamental kinematic gait parameters (i.e., stride length, velocity, foot clearance, and step length), from data collected with the robot's on-board RGB-D sensor and with the instrumented insoles during straight walking and turning tasks. The accuracy of each model was validated against ground-truth data measured by an optical motion capture system with N=10 subjects. Results suggest that the combined use of wearable and robot's sensors yields more accurate gait estimates than either sub-system used independently. Additionally, SVR models are robust to inter-subject variability and type of walking task (i.e., straight walking vs. turning), thereby making it unnecessary to collect subject-specific or task-specific training data for the models. These findings indicate the potential of the synergistic use of autonomous mobile robots and wearable sensors for accurate out-of-the-lab gait analysis.

In this paper, a new adaptive tracking controller with learning ability is proposed for fully-constrained cable-driven parallel robots (CDPRs). For these systems, the necessity of maintaining positive and bounded tensions in all cables while coping with disturbances represents a critical control requirement. To achieve this goal, we propose a control law based on adaptive dynamic programming (ADP), with an actorcritic structure. In the critic part, an artificial neural network (NN) approximates the value function which is to evaluate the system performance; in the action part, the controller’s parameters are tuned online to achieve optimal control performance. Additionally, the anti-windup (AW) technique is combined with the adaptive controller to cope with the input saturation problem. The stability of the closed-loop system with the proposed control algorithm is proved using the Lyapunov method. Numerical simulations show the effectiveness of the proposed controller.

In robot-assisted rehabilitation, assist-as-needed (AAN) controllers have been proposed to promote subjects’ active participation, which is thought to lead to better training outcomes. Most of these AAN controllers require a patient-specific manual tuning of the parameters defining the underlying force-field, which typically results in a tedious and time-consuming process. In this paper, we propose a reinforcement-learning-based impedance controller that actively reshapes the stiffness of the force-field to the subject’s performance, while providing assistance only when needed. This adaptability is made possible by correlating the subject’s most recent performance to the ultimate control objective in real-time. In addition, the proposed controller is built upon action dependent heuristic dynamic programming using the actor-critic structure, and therefore does not require prior knowledge of the system model. The controller is experimentally validated with healthy subjects through a simulated ankle mobilization training session using a powered ankle-foot orthosis.

Adaptive cable-driven parallel robots can adjust the position of one or more pulley blocks to optimize performance within a given workspace. Because of their augmented kinematic redundancy, adaptive systems have several advantages over their traditional counterparts featuring the same numbers of cables. In this paper, we explore the application of adaptive cable-driven robots to cable-suspended camera systems. Performance of the traditional and of the adaptive designs are analyzed, using dexterity and stiffness as performance metrics. Results show superior performance of the adaptive design compared to the traditional system. An illustrative design problem for adaptive cable-suspended camera systems is also presented and solved.

This paper investigates the production and identification of emotional states of a walker using plantar vibro-tactile simulations. In a first experiment, participants were asked to render, according to imagined walking scenarios, five emotions (aggressive, happy, neutral, sad, and tender) by manipulating the parameters of synthetic footstep vibrations simulating various combinations of surface materials and shoes. Results allowed to identify, for the involved emotions and vibration conditions, the mean values and ranges of variation of two parameters, vibration amplitude and temporal distance between consecutive steps. Results were in accordance with those reported in previous studies on real walking, suggesting that the plantar vibro-tactile expression of emotions in walking is independent of the real or imagined motor activity. In a second experiment, participants were asked to identify the emotions portrayed by walking vibrations synthesized by setting the synthesis engine parameters to the mean values found in the first experiment. Results showed that the involved algorithms were successful in conveying the emotional information at a level comparable with previous studies. Results of both experiments revealed strong similarities with those of an analogous study on footstep sounds suggesting that emotionally expressive walking styles are consistently produced and recognized at auditory and plantar vibro-tactile level.

Due to the complexity of the human musculoskeletal system and intra/intersubjects variability, powered exoskeletons are prone to human-robot misalignments. These induce undesired interaction forces that may jeopardize safe operation. Uncompensated inertia of the robotic links also generates spurious interaction forces. Current design approaches to compensate for misalignments rely on the use of auxiliary passive degrees of freedom that unavoidably increase robot inertia, which potentially affects their effectiveness in reducing undesired interaction forces. Assessing the relative impact of misalignment and robot inertia on the wearer can, therefore, provide useful insights on how to improve the effectiveness of such approaches, especially in those situations where the dynamics of the movement are quasi-periodic and, therefore, predictable such as in gait. In this paper, we studied the effects of knee joint misalignments on the wearer's gait, by using a treadmill-based exoskeleton developed by our group, the ALEX II. Knee joint misalignments were purposely introduced by adjusting the mismatch between the length of the robot thigh and that of the human thigh. The amount of robot inertia reflected to the user was adjusted through control. Results evidenced that knee misalignment significantly changes human-robot interaction forces, especially at the thigh interface, and this effect can be attenuated by actively compensating for robot inertia. Misalignments caused by an excessively long robot thigh are less critical than misalignments of equal magnitude deriving from an excessively short robot thigh.

Walking mechanics has been studied for a long time, being essentially simple but nevertheless including quite tricky aspects. During walking, muscular forces are needed to support body weight and accelerate the body, thereby requiring a metabolic demand. In this paper, a new Constant Pushing Force Device (CPFD) is presented. Based on a novel actuation concept, the device is totally passive and is used to apply a constant force to the pelvis of a subject walking on a treadmill. The device is a serial manipulator featuring springs that provide gravity balancing to the device and exert a constant force regardless of the pelvis motion during walking. This is obtained using only two extension springs and no auxiliary links, unlike existing designs. A first experiment was carried out on a healthy subject to experimentally validate the device and assess the effect of the external force on gait kinematics and timing. Results show that the device was capable of exerting an approximately constant pushing force, whose action affected subject's cadence and the motion of the hip and ankle joints.

This paper introduces the overall design of ALEX III, the third generation of Active Leg Exoskeletons developed by our group. ALEX III is the first treadmill-based rehabilitation robot featuring 12 actively controlled degrees of freedom (DOF): 4 at the pelvis and 4 at each leg. As a first application of the device, we present an adaptive controller aimed to improve gait symmetry in hemiparetic subjects. The controller continuously modulates the assistive force applied to the impaired leg, based on the outputs of kernel-based non-linear filters, which learn the movements of the healthy leg. To test the effectiveness of the controller, we induced asymmetry in the gait of three young healthy subjects adding ankle weights (2.3kg). Results on kinematic data showed that gait symmetry was recovered when the controller was active.

Sophia-3 (string-operated planar haptic interface for arm rehabilitation) is a planar cable-driven device with a tilting working plane. It represents the first application of the adaptive cable-driven design paradigm recently introduced by the authors, featuring a moving pulley-block that allows the robot to achieve excellent force capabilities, despite the low number of cables. This study presents the design, kinematics, and control of the device and results of experimental validation on healthy subjects.

ALEX III is a bilateral exoskeleton for gait rehabilitation. It is an evolution of two previous prototypes - ALEX and ALEX II - developed at the University of Delaware. The new robot comprises a support platform and two robotic legs. Its unique characteristic is the possibility to actively control 12 degrees-of-freedom: 4 at the pelvis and 4 for each leg. This paper focuses on the design and fabrication of the robotic leg. Results from early evaluations are presented where the robotic leg is attached to a fixed frame and controlled with a zero-interaction controller.

As our robotics community advances its understanding toward the optimal design of robotic exoskeletons for human gait training, the question we ask in this paper is how the anterior lunge degree of freedom in the robotic exoskeleton affects human gait training. Answering this question requires both novel robotic design and novel protocols for human gait training to characterize this effect. To the best of the authors' knowledge, this is the first study to characterize the effect of an exoskeleton's degrees of freedom on human gait adaptation. We explored this question using the Active Leg EXoskeleton (ALEX) II. The study presented was performed using ALEX II under the following two configurations: 1) locking the anterior/posterior translation in the exoskeleton, while allowing other degrees-of-freedom (labeled as locked mode) and 2) keeping the anterior/posterior degree of freedom unlocked (labeled as unlocked mode). Healthy subjects walked at self-selected speeds on a treadmill and were trained to walk with a new gait template, scaled down from their normal template. While both groups showed adaptation and retention over a 26-min period following training, the unlocked group showed better performance in terms of adaptation and retention compared with the locked group.

This work deals with the kinematic and dynamic modeling of a 3-DOF, under-actuated, pendulum-like manipulator and its control system. The cable-based device is capable of completing point-to-point planar motions, driving the end-effector from a starting pose to a goal pose, by means of two actuators only. The device relies on parametric excitation to control the oscillations of the variable-length pendulum. Unlike a previous work, the dynamic model introduced here is consistent with the assumption of cable-based device. Several control strategies are compared through numerical simulations.